The hype is extreme right now and everyone is still trying to figure out how to use the tools. People who are further along the bleeding edge are trying to tear down all the process that we used to have to further improve velocity. After people go all the way to "dark factories" most companies will realize that they don't actually have any good ideas for what to build, and honestly never did. They've been coasting for years, someone else can replicate their product now and its just a race to the bottom. At that point, token budgets are going to collapse.

I'm staunchly pro-AI as a technology, but I do think the bubble is going to pop in the next year or two just because the business value won't materialize for most companies fast enough.